We present a multidisciplinary workflow for analysis of digital rock samples and recommendation of Enhanced Oil Recovery (EOR) materials using Artificial Intelligence, Digital Rock Physics and laboratory experiments based on scalable computational and experimental platforms, such as Cloud Computing and CMOS. This multistep workflow implements a progressive screening methodology that reduces the number of experimental trials by pre-evaluating candidates using data-driven and physics-driven computational methods. A knowledge base containing well-logs, fluid properties and EOR material data is populated from public (patents and papers) and private (production-level, simulations and laboratory) sources. Artificial Intelligence algorithms screen EOR materials for a given reservoir scenario and provide recommendations based on efficiency and suitability. Oil recovery enhancement effects are simulated at pore-scale for the most promising candidates based on a three-dimensional representation of the rock, thus reducing the candidate list. The remaining EOR materials under consideration are submitted to fluid flow experiments in the laboratory using a customised device platform (rock-on-a-chip) that reproduces the pore-scale structure of the relevant reservoir rock. The experimental results further quantify the EOR efficiency of each material and allows selecting the best candidate. In this work we present examples of application of such workflow.